ANALYSIS OF PALM KERNEL CAKE BY NEAR INFRARED REFLECTANCE TECHNOLOGY

NORLINDAWATI A.P.*, SHARIFF S., NOORMAH MIW A., SABARIAH B., OMAR R., SAMIJAH A., SUPIE J. AND JAMNAH O.

NORLINDAWATI A.P.*, SHARIFF S., NOORMAH MIW A., SABARIAH B., OMAR R., SAMIJAH A., SUPIE J. AND JAMNAH O.

The rapid method for predicting palm kernel cake quality with near infrared reflectance spectroscopy was investigated. Chemical tests for moisture, ash, crude protein, crude fat, crude fibre, total digestible nutrients (TDN), nitrogen free extract (NFE), metabolisable energy (ME), calcium and phosphorous were time consuming and involve high cost. This technology can save considerable time by testing all the parameters simultaneously; however accurate calibration of the equipment is essential. Near infrared (NIR) partial least square (PLS) regression models for determination of several palm kernel cake quality parameters were developed from NIRFlex Model N-500 (Buchi). In general, reliable prediction results were obtained for the TDN (SEP = 0.85 r2 = 0.99), NFE (SEP = 0.42 r2 = 0.97) and crude protein (SEP = 0.57 r2 = 0.98) PLS regression models.